Research Of Visualization Of Underwater Acoustic Data

Acoustic technology has been applied more and more widely in the field of underwater detection and utilization and exploitation of ocean resources. As the main way of underwater exploration, underwater acoustics imaging technology has been rapidly developed. The underwater acoustic data visualization technology of reconstructing the data being collected by underwater sonar with three-dimensional visual effects is the current research focus. Because of the complex underwater environment, the propagation of sound wave can be easily affected by a variety of interferences. Thus in the generation and transmission process of the data collected by sonar will inevitably be affected by noises, this made the background information more complex. What’s more, acoustic data itself has low contract, we will not get a good visual effect if we reconstruct the data before any processes. So it is necessary for us to reduce the noise contained in original data. And segment the data to distinguish different background information in order to observe each part of the underwater environment more conveniently.Based on the research of traditional noise reduction algorithms, classical segmentation algorithms and the characteristics of the three-dimensional underwater acoustic data, this thesis presents a complete solution for underwater acoustic data visualization. That is, firstly reduce the noise contained in the original acoustic data, then segment the de-noised data, finally use the visualization toolkit VTK to verify the treatment efficiency.This thesis presents a 3D underwater acoustic data de-nosing algorithm for visualization firstly. This algorithm analyzes characteristics of the composite noise model, and considers the additive noise and multiplicative noise respectively. With combining spatial domain and transform domain, this paper proposes a layered de-nosing method by using fuzzy medium filter and soft threshold filter based on wavelet transform with considering the correlation of wavelet coefficients respectively, and 3D neighborhood information of acoustic data being as far as possibly used in the process. Experimental results show that the proposed algorithm reduces the noise in the original data at the same time improving the data continuity and contrast, and enhances the visual effect of 3D underwater acoustic data.This thesis presents a fusion segmentation algorithm of 3D underwater acoustic data secondly. This algorithm is combined with fuzzy c-means, mathematical morphology and an improved OTSU threshold algorithm. As the water, seafloor and their transition with different acoustic characteristics, the data is coarsely divided into three parts firstly, each part is then further segmented finely according to their different characteristics and get different segmentation templates. Then remove unwanted background information in accordance with the templates. Experimental results show that the proposed algorithm improves the detection accuracy of underwater targets and the performance of underwater sonar data.Finally, this thesis uses the visualization toolkit VTK to achieve visualization of the processed 3D underwater acoustic data by the ray casting algorithm of direct volume rendering. After that we can not only observe every part of the data independently, but also the whole assembled data. And the result shows the effectiveness of the data processing of de-noising and segmentation.